People with severe mental illness (SMI) die 25 years earlier, mostly due to cardiovascular disease and diabetes, which are directly related to obesity. Obesity is a leading cause of preventable death in the US, second only to smoking. There is an urgent need for the development of interventions for the growing problem of obesity as well as related morbidity and mortality in this vulnerable population. Poor knowledge about nutrition, lack of access to healthy foods, and the increased costs of healthy foods contribute to increased prevalence of obesity among SMI patients. Regardless, the most important contributor to the obesity epidemic in this population is antipsychotic-induced weight gain. Behavioral interventions are necessary to fight obesity, but are insufficient, as patients continue to need their psychiatric medications, which in turn increase their appetite and make the task of losing weight extremely difficult. We have developed a working model of antipsychotic-induced weight gain. We hypothesized that antipsychotic medications attenuate the rewarding effects of food, thus causing SMI patients to eat more to get the same level of reward. To prove the concept, we completed a pilot study using naltrexone, a medication which blocks the effects of endorphins, the brain proteins which mediate the reward. If the additional reward is not gained with eating more, patients should stop doing so. Indeed, the pilot study showed that naltrexone stopped the weight gain trend in a group of patients taking antipsychotics and, in fact, produced significant weight loss for non- diabetic patients, while the placebo group continued to gain weight during the trial. This is an indirect evidence to support our hypothesis, as naltrexone does not produce weight loss in standard obesity treatment. Based on this evidence, naltrexone may be a useful, easy to use and cost effective medication to counteract antipsychotic-induced weight gain. In this application, we propose a large double blind, randomized clinical trial utilizing two doses of naltrexone compared to placebo over a period of one year (52 weeks) to counteract antipsychotic induced weight gain and obesity among a group of patients with SMI.